Natural History, #c. J99 



and smoothly truncated perpendicularly to these sides, was some- 

 where about half a degree in width, and in length about eight 

 degrees, though varying in both particulars. They were not radii 

 from the north, but parallel to each other, running from a little east 

 of north, their lower extremities being about 20 from the horizon. 

 The bearings of these columns differed much from that of the arch 

 before mentioned. They soon became merry dancers, were then 

 bent rapidly, and continued nimbly playing into curves of small 

 circles, sometimes looking as if gracefully folded or twisted like the 

 most delicate gauze. The order of the colours, whether the same 

 with those of the rainbow, the same in all the columns, and the red 

 towards east or west, was not noted at the time. 



The iris-like appearance continued only a few minutes. The sky 

 soon became pure ; all clouds disappeared, and long bright streamers 

 shot up from the north to the zenith ; some of them continued half 

 a minute, and were occasionally tinged red or yellow. After ihese 

 streamers, feeble lights from the north rose like faintly luminous 

 puffs of smoke, then occurred the common lights, and in about fif- 

 teen or twenty minutes from the first all was over. 



The moon, nearly full, rose about eight o'clock, and shone the 

 whole time, but nothing justifies the reference of the colours to her 

 light, on the contrary many circumstances oppose it. The air had 

 towards evening become rapidly cool, and the atmosphere during 

 the day had been smoky from the burning wood, but these are 

 considered to have no relation to the colours *. 



23. INFLUENCE OF THE AGE OF PARENTS ON THE SEX OF 

 CHILDREN. 



The following results are extracted from a letter written by 

 Professor Hosacker to the Editor of the Medical Gazette of 

 Inspruck : 



i. In those marriages where the mother is older than the father 

 the number of boys is to that of girls as 90.6 to 100. ii. The 

 parents being of the same age, the boys were to girls as 90 to 100. 

 iii. The father being from three to six years older than the mother, 

 the proportion of boys to girls was as 103.4 to 100. iv. The father 

 being from six to nine years older than the mother, the boys were to 

 the girls as 124.7 to 100. v. The father being from nine to twelve 

 years older than the mother, the male children were to the female 

 as 123.7 to 100. vi. The father being eighteen years or more 

 older than the mother, the boys were to the girls as 200 to lOOf. 



24. PRECAUTIONS IN THE PLANTING OF POTATOES. 



It would appear from experiments made in Holland, that when pota- 

 toes are planted, the germs of which are developed, as happens 

 occasionally in late operations, or after mild winters, that the produce 



* SilUman's Joum.,xviii., 72. f Bib. Univ., 1830, 456. 



